Questions & answers. Web Analytics - GA4, GTM, e-commerce, GDPR and data quality.

GA4 is built on an event model instead of UA views, sessions and hits. It collects data in one property for web and apps, handles cookie loss better (behavioural modelling) and offers BigQuery export. Universal Analytics stopped collecting data in July 2023 - new projects should start in GA4.

Every user interaction - page view, click, purchase, form submit - is an event with a name and optional parameters (e.g. item_id, value, currency). GA4 has no separate hit types like UA; you configure which events are key business conversions.

A conversion is an event marked in GA4 admin as a key event (formerly conversion). It can be purchase, generate_lead, sign_up or any custom event. Without correct conversion marking, attribution reports and Google Ads optimisation rely on the wrong signals.

Common causes: different time zones, cancelled orders counted in the shop but not in GA4, missing purchase event, duplicate transactions, ad blockers, cookie consent refusal, different session definitions and processing delay (GA4 can show 24-48h lag in some reports).

Consent Mode v2 sends Google information about user consent (analytics, ads, personalisation) and enables modelling of missing data when users refuse. Since March 2024 it is required for some Google Ads features in the EEA. Connect your cookie banner with GTM/GA4 - otherwise you lose data and risk non-compliance with regulator expectations. First choose Basic or Advanced in the comparison guide, then validate your setup with the audit checklist.

GTM is a container for tags (GA4, Facebook Pixel, ads, etc.) managed in a UI without deploying code for every change. It provides triggers (click, scroll, form), variables and versioning. It is used to implement and test tracking faster and more safely.

gtag.js on the page sends data straight to GA4 - simple, but every change needs a code deploy. GTM loads one snippet and you configure tags, conditions and variables in the panel. With several marketing tools, GTM is practically standard.

The dataLayer is a JavaScript array where the site passes event and data information to GTM (e.g. product, cart value, category). GTM reads it and sends data to GA4 or other tools. A correct dataLayer is the foundation of precise e-commerce and conversion tracking.

Use GTM Preview mode - connect the site to the container and see which tags fired with which parameters. Complement with GA4 DebugView and browser dev tools. Do not publish the container without testing on an environment close to production.

Tags run on your server (e.g. Google Cloud Run), not in the browser. It reduces page load, makes AdBlock bypass harder and gives more control over data before sending to GA4 or ads. It requires server setup, first-party domain and infrastructure maintenance.

Link your Google Ads account to the GA4 property (Admin → Google product links). You can import conversions from GA4 (key events) or measure them with a separate Google Ads Conversion Tracking tag in GTM. For e-commerce, the purchase tag should send the same parameters as GA4 (value, currency, transaction_id). Check taxonomy in the event planner, GTM checklist and Meta Pixel vs Google Ads guide.

Meta Pixel is a browser tag (via GTM or on-page code) sending events to Meta Ads (PageView, Purchase, Lead, etc.). CAPI sends the same events from your server - more resilient to AdBlock, ITP and cookie limits. A solid 2026 setup uses pixel + CAPI with event_id deduplication. See the Meta Pixel vs Google Ads guide, server-side services and the SS solutions catalog.

Common causes: different attribution windows and conversion definitions, ad_storage consent refusal (Consent Mode), duplicate or missing purchase tags, importing only some GA4 key events, reporting lag between Ads and GA4, and modeled conversions in Advanced Consent Mode in one system but not the other. Compare by transaction_id and verify the Ads tag fires after marketing consent.

Advertising tags usually require marketing consent (ad_storage, ad_user_data in Consent Mode v2). In GTM use blocking triggers or consent checks - pixel and Ads must not fire before banner acceptance. Choose a mode in the Consent Mode selector, validate with the audit checklist and test refusal in GTM Preview.

Key metrics: purchase revenue, transaction count, average order value (AOV), conversion rate, add to cart, cart abandonment and engaged sessions. Combine with campaign attribution and product reports (item-scoped dimensions).

Minimum full funnel: view_item (or view_item_list), add_to_cart, begin_checkout, add_shipping_info, add_payment_info, purchase. For product lists - select_item. Each event should have an items array with item_id, item_name, price, quantity plus currency and value where GA4 requires them.

Common causes: event fired before payment confirmation, no transaction_id deduplication, wrong dataLayer on thank-you page, users without cookie consent, different time zones or test orders filtered on only one side. Verify in DebugView and compare by transaction_id.

Yes, if they support marketing decisions - remarketing, product list optimisation, A/B tests. Mark as conversions only events that truly match business goals; too many conversions blur reports and make campaign optimisation harder.

In practice yes - GA4 uses identifiers and may transfer data outside the EEA. Under GDPR and EU regulator guidance, analytics and advertising tags should run after informed consent (CMP banner), unless you have another legal basis confirmed with legal counsel.

It depends on configuration: GA4 alone in analytics mode is often classified as analytics (consent required in many jurisdictions). Linking with Google Ads, remarketing or User-ID usually requires marketing consent. Classification should follow your privacy policy and legal advice - not the tool settings alone.

Your CMP (Cookiebot, OneTrust, custom) should update consent state in the dataLayer or fire Consent Mode (ad_storage, analytics_storage, etc.). In GTM, GA4 and ad tags fire on consent triggers, not immediately on page load. Test both refuse and accept in Preview.

When you need raw data without sampling limits, custom SQL reports, joining GA4 with CRM/ERP, long retention (GA4 UI has historical limits) or historical audit. Daily export is free on the GA4 side; you pay for BigQuery storage and queries.

No - BigQuery is the analytical layer for analysts and data engineers. GA4 UI still serves quick insights, explorations and daily monitoring. Choose BigQuery when standard reports are not enough or you need full control over metric definitions.

Cost depends on event volume and query frequency. For many SMB sites it fits in Google Cloud free tier (storage + first TiB of queries). At high traffic, plan partitioning, table retention and SQL query optimisation.

Microsoft Clarity is a free behavioural tool: heatmaps, session recordings (session replay), rage clicks and dead clicks. It does not replace GA4 - it adds the “why users don't convert” layer. GA4 answers how many and from where; Clarity shows on-page behaviour. See the <a href="/en/microsoft-clarity-checklist/">Clarity implementation checklist</a>.

In practice yes - Clarity uses cookies and collects behavioural data, so in the EU the tag should fire after the user's consent in the CMP (analytics category). Don't load Clarity before the cookie banner. Verify your setup in the <a href="/en/microsoft-clarity-checklist/">Clarity checklist</a> and the <a href="/en/consent-mode-checklist/">Consent Mode v2 checklist</a>.

Create a project at clarity.microsoft.com, copy the Project ID and add a tag in GTM with a trigger that fires after analytics consent. Avoid a double install (hardcode + GTM). Enable form-field masking in the project settings. Step by step: see the <a href="/en/microsoft-clarity-checklist/">Clarity checklist</a>.

Clarity is free and enough for heatmaps and replay at typical site traffic. Hotjar adds surveys, feedback and A/B testing in paid plans. If you only need UX diagnostics alongside GA4 - start with Clarity.

Yes - in the Clarity panel you can connect the project with GA4 and filter session recordings by GA4 segments. This makes analysis easier: you see a conversion drop in GA4 and immediately watch sessions from that segment in Clarity.

Piwik PRO is an analytics platform focused on privacy and EU data hosting. Consider it with stricter compliance needs (public sector, healthcare), when you need full control over the data server or want less dependence on the Google ecosystem.

Yes - many companies collect into both via GTM (separate tags, shared dataLayer). Important: dual loading should not slow the site excessively, and the cookie banner must cover both tools. Metric definitions will still differ between systems.

After UA migration, before increasing ad spend, when numbers do not match shop or CRM, key events are missing or the team does not trust reports. An audit covers tags, dataLayer, conversions, CMP consent and Measurement Plan alignment - ending with a prioritised fix list.

A document linking business goals to measurable events: which conversions, which parameters, which reports, which data sources. Without it, GA4 implementations often collect “everything” but answer no leadership or marketing questions.

Simple corporate site: a few to several business days. Shop with full e-commerce, GTM server-side, BigQuery and dashboards: usually several weeks. Timeline depends on domains, integrations, client team availability and compliance requirements.

Use DebugView and realtime reports, compare transactions with backend by transaction_id, verify Consent Mode in GTM Preview and that ecommerce events have items, currency and value. Audit checklists help structure the review but do not replace production tag analysis and historical data review.